A team of researchers from San Diego applied a machine learning algorithm to genetic screening. The goal was to speed up and improve the diagnosis of some rare genetic diseases. In this way it will be easier to intervene quickly, especially in the case of babies in intensive care.
The study was conducted in collaboration with Illumina. Dr. Kingsmore's team used genetic sequencing on infants and children in intensive care. All patients were in severe condition, requiring immediate intervention. To intervene in the right way, however, a diagnosis as precise as possible was needed. So here comes the algorithm. The new technology reduces the need for human intervention in the analysis of genetic data. This reduces the time and cost of the intervention, giving reliable results over 19 hours. All from a simple blood sample, which contains all the genetic material necessary for total sequencing.
The key components of the technology come from Illumina, a company specializing in prenatal screening tests and genetic tests. These work with machine learning systems, which learn how to interpret the data available to them better and better. To make this possible, the researchers entered data from the medical literature into the system.
The algorithm evaluates them and compares them with the data obtained from sequencing, providing the possible answers. It takes about 5 minutes. The new algorithm does not replace human experts, but simplifies the work. Widely applied, it could save the lives of thousands of children with genetic diseases.